Next Word Prediction App – Final Project Presentation

Nishath Anjum P

📌 Slide 1: Project Overview

This project demonstrates a Next Word Prediction application built using R, Shiny, and a lightweight text prediction model.

The app: - Takes a phrase as input - Predicts the next word - Runs interactively on shinyapps.io

This submission fulfills the data product requirement for the Coursera Final Project.

⚙️ Slide 2: How the Prediction Model Works

Method Used: Word-pair matching (bigram lookup)

Steps: - Extract last word from user input - Match it against known word pairs - Return most likely next word - If no match → return “unknown”

Libraries used: - stringr for tokenizing - dplyr for filtering - shiny for app interface

💻 Slide 3: App Demo & User Experience

How the app works: 1. User enters a phrase → e.g., “machine” 2. Click “Predict” 3. App returns: “learning”

🔗 App Link: (replace with your URL) https://nishath.shinyapps.io/Nextword-app/

This allows anyone to interact with the predictor online.

🚀 Slide 4: Deployment & Technology

Tools Used: - RStudio - Shiny - shinyapps.io - Quarto for slides - RPubs for hosting

Deployment: - Run App → Publish - Hosted via shinyapps.io with public URL access

🎯 Slide 5: Conclusion & Next Steps

This project demonstrates: - Working prediction model - Live deployment - Interactive user interface - Full reproducibility in R

Future improvements: - Add n-gram model - Use real NLP training data - Improve prediction accuracy

Thank you for reviewing!